Educators are the driving force behind who is picked for special school services, such as gifted-and-talented programs. It turns out that racial bias may play a stronger role in the selection process than previously thought.

The major deciding factor in how a teacher classifies two students showing identical signs of giftedness or disability may not be their abilities.

It's likely to be a student's race or ethnicity, according to a new study from New York University.

Educators have known for years that Black and Latino students are less likely than their white or Asian-American classmates to be assigned to gifted-and-talented programs.

How different are their chances?

It turns out that they're much lower:

• Black students are 66 percent less likely to be assigned to a gifted-and-talented program than their white peers.
• Latino students are 47 percent less likely.

Even among students with high scores on standardized assessment tests, the disparity still exists between white and Black children.

White students with high scores in math or reading were more than twice as likely as equally high-achieving Black students to be selected for academically advanced programs.

This raises the question: Do gifted programs actually work?

In other words, do such programs serve students who exceed expectations for their grade level, or do they simply perpetuate advantages that are already baked into the system because of racial bias?

It's hard to say.

It does seem that race affects a child's chances of being identified as academically advanced, the New York University researchers found in the study, which will be published in an upcoming issue of the journal Social Science Research.

It's possible that there is a direct link between who is referred to special programs and racial bias among teachers.

Researchers led by NYU professor Rachel Fish took 70 third- and fourth-grade teachers from a large, northeastern city and assigned them to read profiles of fictional male students who each showed signs of academic difficulties, behavioral challenges, or academic giftedness.

The profiles described identical behaviors, with one difference: the name of the student, which was changed to signal different racial identities.

• Teachers who participated in the study had a higher chance of seeing academic gifts in profiles assigned to the name Jacob and were less likely to assign an identical profile with the name Carlos or Demetrius to a gifted-and-talented program.
• Academic and behavior problems in the profiles with the whitest names were often seen as "medicalized problems to fix" by teachers.
• The exact same behaviors were seen as ordinary when the profile name was either Carlos or Demetrius. This indicated that "low academic performance is normal for [students of color] and not a problem to remediate."

"The issue is that racism affects all of us, and teachers are in positions of power," study author Fish said.

This isn't the first time that teachers have been implicated as part of the problem.

A previous study indicated that Black students were three times less likely to be assigned to gifted reading programs when they were taught by a non-Black teacher, versus when they were taught by a Black educator.

Researchers warn that neither of these studies should be taken to mean that all teachers are racist, which would be a dramatic over-generalization. Rather, the studies suggest that implicit bias — or bias that occurs below the conscious level — is one of many potential problems that are enabled by larger institutional factors.

Interested in learning more about implicit bias? You can take your own implicit association test here.